frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

U.S. approves $2.1B sale of Patriot interceptors to Ukraine

https://www.nbcnews.com/world/ukraine/us-approves-21-billion-sale-patriot-interceptors-ukraine-rc...
1•paaloeye•2m ago•0 comments

Fair Use Benefits Everyone

https://openai.com/index/fair-use-benefits-everyone/
2•pentagrama•5m ago•0 comments

The 'Moylan Arrow' Is One of Ford's Best Contributions to Car Design

https://www.thedrive.com/news/the-moylan-arrow-is-one-of-fords-best-contributions-to-car-design
3•thunderbong•8m ago•0 comments

10.0: CVE-2026-112816 – Canonical LXD – Improper cert validation — no emudhra

https://twitter.com/cvedotnow/status/2116213963872518418
1•paaloeye•9m ago•0 comments

Ask HN: Are most of you using HN Replies? Or what else do you use?

1•Gooblebrai•10m ago•0 comments

Show HN: Free Unlimited Virtual Monitors+Watch Claude Work on Them+Community

https://multispace.dev/
1•DanielSystem•22m ago•4 comments

The United Kingdom is committing an act of strategic self harm

https://blog.rodmena.com/entry/event_horizon_machine_standards
4•rodmena•23m ago•3 comments

Federal Bureau of Prisons employees serve as election monitors at polling sites

https://www.corrections1.com/federal-prison/bop-employees-to-serve-as-election-monitors-at-pollin...
10•MadrasTh0rn•27m ago•2 comments

AI changed what I can build

https://tanner.place/blog/how-ai-changed-what-i-can-build
2•vrennat•29m ago•1 comments

Zellic's and V12's Work on x86 Binary Formal Verification

https://v12.sh/blog/x86-formal-verification-intro
1•drdrey•29m ago•0 comments

Anthropic to receive high 'junk' rating from credit agencies following IPO

https://techcrunch.com/2026/10/30/anthropic-expected-to-receive-high-junk-rating-from-credit-agen...
5•jnord•31m ago•0 comments

Worst Halloween candies, according to the data

https://flowingdata.com/2026/10/30/worst-halloween-candies-according-to-the-data/
2•surprisetalk•31m ago•0 comments

Analysts find looming launch crisis is "more severe" than anticipated

https://arstechnica.com/space/2026/10/analysts-find-looming-launch-crisis-is-more-severe-than-ant...
4•jnord•31m ago•0 comments

Reddit shares plunge as Google's AI search squeezes traffic

https://www.reuters.com/business/reddit-pinterest-shares-tumble-weak-user-revenue-growth-2026-10-30/
1•tartoran•36m ago•1 comments

Tell HNN: Pets.com was *not* owned today

2•y1n0•37m ago•0 comments

Hands on the Wheel

https://ryanbigg.com/2026/10/hands-on-the-wheel
2•ryanbigg•38m ago•0 comments

The N64's GPU Is a Technical Marvel [video]

https://www.youtube.com/watch?v=WEoRgQH08MA
3•zdw•41m ago•0 comments

The Thaura Constitution

https://thaura.ai/constitution
2•conferza•45m ago•0 comments

Somoene bets $10M on Reddit stock anad loses half of it

https://old.reddit.com/r/wallstreetbets/s/LWYApzWCQH
1•paulpauper•45m ago•0 comments

The Ground Beneath Them Was Dry

https://lagomor.ph/2026/10/the-ground-beneath-them-was-dry/
2•jfil•46m ago•0 comments

Sndstation will transform cheap gaming consoles into music handhelds

https://cdm.link/sndstation-linux-distro-for-game-handhelds/
2•Blackthorn•46m ago•0 comments

The Commanding Heights

https://commanding-heights.com/
1•emersonmacro•50m ago•0 comments

$20 plans, $1 of Sol costs 7x what $1 of Opus does. A task, only 1.3x

https://twitter.com/AdamHoltererer/status/2115879037893534195
1•marcellosso•50m ago•0 comments

Assure – your AI agent says "done." This checks if it's lying

https://github.com/mgasparin67-cmd/Agent-verification_Assure
2•spaghetti0•51m ago•0 comments

Ukrainian arms manufacturer Fire Point has overcome Starlink restrictions

https://www.reuters.com/business/aerospace-defense/ukrainian-arms-maker-fire-point-says-its-syste...
6•uxhacker•53m ago•0 comments

AI's Bill Is Due – Intelligence or Brand? – Cold Logic – EP01

https://www.youtube.com/undefined
1•LebToki•58m ago•0 comments

Zinc creep in cable spelter sockets brought down the Arecibo telescope in 2020

https://maketecheasier.com/arecibo-zinc-creep-collapse/
3•ohjeez•58m ago•0 comments

The reasons embedded projects get into trouble (2002)

https://www.ganssle.com/articles/jackstoptenlist.htm
2•ohjeez•1h ago•0 comments

Show HN: Fastbrowse, Browser agent that picks clicks with a System One model

https://www.fastbrowse.ai/
3•furqanrydhan•1h ago•2 comments

Defining AI

https://ali-alkhatib.com/blog/defining-ai
2•bediger4000•1h ago•0 comments
Open in hackernews

Show HN: Fastbrowse, Browser agent that picks clicks with a System One model

https://www.fastbrowse.ai/
3•furqanrydhan•1h ago
Hi HN, I'm Furqan founder of Nebula. We've spent a lot of time building, using and breaking browser agents. The main issues we ran into were costs and speed. So we built our own and open sourced it.

It's pretty simple, give it a task and it drives Chrome faster and cheaper than the alternatives.

At each step it lists the page's elements instead of screenshots. This means it can't click on buttons that aren't there.

We use System One models (currently Jev), which keeps costs extremely low, around 11x lower on average. If Jev does fail, we fall back to a cheap LLM like Gemini Flash Lite or Deepseek v4.1.

We ran our 46-task suite against Browser Use's hosted agent at its default settings. Each task ran three times, both on their cloud browsers. fastbrowse cost $0.030 a run, theirs cost $0.318. We passed 137 of 138 runs, and they passed 133.

To set up all you need is uv, Chrome, and a TypeSafe or OpenRouter key, both work:

uvx fastbrowse "What is the title of the top story right now?" --start https://news.ycombinator.com/ --local

Would love to hear feedback on it and how it's working for you guys.

https://github.com/agent-labs-dev/fastbrowse

Comments

lando_cal•55m ago
interesting, want to see how it performs. been using a browser use for a while so curious to see
szaqadir•52m ago
very cool! how does it handle sign in tasks?